SUMMARY OF WORK The new 3D model of stochastic calcium release in sino-atrial node cells (SANC) has been elaborated and revised to provide extensive imaging and movie output, and to allow simulations of indefinite length in order to study rhythm effects. An attempt was made to base the model geometry directly on observed immunofluorescence images of RyR distribution, but it was found that the information in conventional light microscopic images has insufficient resolution to define the local pathways by which CICR spreads between neighboring couplons. It will be necessary to obtain super-resolution microscopy of these cells. For this reason we have established a collaboration with a microscopy group in NIBIB. Preliminary collaborative studies have confirmed the model result that calcium release events are confined to a thin layer beneath the sarcolemma, which would not be demonstrated with standard confocal microscopy. The model has been extensively exercised and a large paper has been published in Journal of General Physiology. Further studies have been directed to a pathological regime discovered by modeling, in which the regulatory effects of the adrenergic system on heart rate are reversed when RyR sensitivity is too high. We have begun exploring this regime -- of possible clinical significance in heart failure and genetic abnormalities of RyR and CASQ -- using caffeine to modulate the RyR. Initial results show that, in order to understand arrhythmic beating of SANC in the coupled-clock regime, it will be necessary to improve on the underlying electrophysiological model of these cells that has been widely accepted. We are also using the 3D model to study the mechanism of heart rate variability at the single-cell level, which requires simulations lasting for days.